In the present work, we investigated a practical use of the Topological Fragment Spectra based support vector machines (SVMs) for drug discovery by virtual screening of antihypertensive drugs. We employed two MDDR databases which were released in different years (2001 and 2003). First, we developed a classification model, which consists of collective SVMs, to identify antihypertensive drugs. We used a structure database, MDDR 2001 to develop and validate the SVM models. The data involved 9503 antihypertensive drugs and the negatives of 66521 compounds. The obtained SVM models gave good performance in the learning, in the validation and in the prediction. To evaluate the ability of the classifiers in practical use, we also prepared an external test set that was derived from the set of compounds in MDDR 2003, but not from the set in 2001 (relative complement). The external test set consisted of 19387 compounds, 396 antihypertensive drugs and 18991 negative compounds. The computational trial of the prediction with the external test set successfully identified 158 compounds of the 369 antihypertensive drugs. The results indicate that the present approach can be useful in virtual screening for drug discovery.
We present some practical studies of drawing the graphs of the radial functions and the distribution functions for the hydrogen-like atom and of drawing the three-dimensional contour plots of H2+ molecular-orbitals and hybrid orbitals by using Microsoft Excel in university courses. Some of the three-dimensional contour representations thus obtained are not consistent with figures given in various textbooks of quantum chemistry, and the usual explanations are liable to cause misunderstandings.
The development of polydimethylsiloxane-based microfluidic devices with integrated controlling components has led to the translation of many bench-top biological processes to microdevice platforms. While many biomedical scientists recognize the applicability of microdevices to their research, their lack of engineering background and training in computer-aided design (CAD) makes them unable to design the desired microchips using AutoCAD software, which is the universal first step for chip fabrication. In this work, we used the .NET platform to develop a user-friendly AutoCAD add-on feature that includes the ability to create channels, loops, valves, and punches, as well as channel to loop connections, valve to channel connections, valve to loop connections, channel to channel connections, and punch to channel connections. This add-on feature greatly simplifies the process of AutoCAD drawing, which allows people not familiar with AutoCAD to easily learn to use the add-on. For testing purposes, we used the add-on to draw a series of two-layer microchips designed to synthesize luciferase using a coupled transcription-translation system with different capacities. The AutoCAD files for the microfluidic designs were then sent to the Stanford Microfluidics Foundry for chip fabrication. Our results show that the quantity of luciferase produced on the chip is proportional to the capacity of the reaction loop.
Recently, significant improvements have been shown in proteomics, an analytical technology recognized as a method to discover biomarkers. Several methods for proteome differential display have been developed to discover biomarkers, and these are based on labeling methods such as ICAT™ reagent. These labeling methods are not easily able to detect biomarkers in complex biological samples, and they have several issues such as the limitation of the modification site and the production of artificial errors. We therefore developed a computational quantitative proteomic analysis system i-OPAL (internal standard guided Optimal Profile ALignment) based on LC-MS measurements for biomarker discovery. This system can provide extensive quantitative analysis because it does not involve any artificial labels that might cause sampling bias and limitation of quantifiable proteins.
Using molecular mechanics with amber potential, we investigated the reaction mechanism of an iodine precipitation reaction induced solute freezing process or the in macro ice solid phase of iodide ionic acidic solution. Though a redox reaction such as iodine precipitation in iodide ionic acidic solution is usually difficult to proceed even in high temperature, the reaction: I- →I • +e- and I • +I • →I2 were processed in solute freezing process or in macro ice solid phase. As initial geometries, we took structures of I2(H2O)30 cluster where waters are surrounding two iodine atoms. Iodine atoms were located outside of water clusters in all geometries obtained by geometry optimization with Amber potential. Unfortunately, geometry optimizations by PM3 and HF/Lanl2DZ were incomplete because iodine atoms were located outside and dissociated from the water cluster. In low temperature, the contribution of enthalpy for system stabilization is bigger than that of entropy. Water crystallization is enhanced by maximization of the number of hydrogen bonding in the solution. Therefore, precipitation of I2 is realized.
In order to analyze of crystallization processes, a tool with spreadsheet software such as Microsoft Excel has been developed to calculate operation parameters for crystallization based on the design theory of an industrial crystallizer. Using this tool, we calculated the nucleation rate and the linear growth rate that are the basis for analysis of crystallization phenomena, and we analyzed the variation of monthly aging and the start-up of the real industrial crystallization processes of two products. As a result, it was observed that product X (sulfated compound) had been produced with nearly perfect stability. On the other hand, the phenomena in the crystallizer at the start-up of product Y (phosphate compound) could be understood. This tool is effective for accumulation and analysis of operation data in the crystallization processes.